Automatic AI Writing Assistant — By Itself or after Dictation
Speak.Get polished text.
Save 5–40 hours a month per person by occasionally speaking instead of typing.
- No credit card required
- Email, DMs, CRM, AI chats, … — and any other app/site
- AI writing assistant without leaving the app you’re writing in
Hi Anna,
Could you please prepare your top three recommended initiatives for improving productivity through AI for our meeting on Friday?
Best regards,
Jonas
Everything you say comes out automatically well-written, or exactly as you say it, depending on the style.
Can be personalized to sound like you or your company.
The days of being disappointed by speech to text are over. WritingAnt is built for Europeans and gives you the option of more AI power than many similar products, for truly well-written and corrected text.
- The AI behind it improved, and so did how it is tuned for good writing.WritingAnt optimizes AI models for writing, with quality checks and extra rounds of improvement where needed.
- You choose the writing style for the situation:
- Instant: usually under half a second.
- High quality: usually 1–3 seconds, for text people will read, where the writing matters more than milliseconds.
- Developed in Switzerland, with privacy as a priority.The interface offers 20 European languages, and your text follows regional spelling, such as British or American English, Swiss “ss” instead of “ß”, and similar variants in other languages.
Works everywhere you type
Wherever you’d normally type, just speak — WritingAnt writes it for you.
No switching programs. No copy-paste.
How it works
Both are included. Pick a tab to see how it works.
Write something new — by voice
Place your cursor in a text field and double-press Command to start recording.
Chat, email, CRM, docs, AI chats — any app.
Talk for up to 20 minutes. No auto-stop, no stress.
The style is picked automatically from the app you’re in: formal in Outlook, relaxed in WhatsApp.
Want another tone? Pick any of the 20+ styles, or customize your own.
One download, both features — install once on your Mac or Windows computer, then it’s ready in every app.
Total since the European launch
WritingAnt speech-to-text sessions, and counting
Write 3× faster by speaking
Watch the same message being written by speaking compared to a keyboard.
Nine minutes to type. Three to say it.
Users tell us the same thing: the task gets done now instead of later. You answer while it is still fresh. You write the full reply, not the lazy one. And it goes out in three minutes.
Ten of those in a day and it shows. You become the person who answers first — and answers properly.
Personalization
Make it sound like you. Or like your company.
The 20+ built-in writing styles work out of the box. You can also define your own (describe the style, who is writing and for which audience), or teach it with your corrections as you go.
Learn more about personalizationPeople notice who writes well — and replies fast.
A clear, fast answer tells people they matter to you. And how you make people feel is how you’re remembered.
Free for macOS and Windows. No credit card.
One speech-to-text tool.Every app you write in.
Speech-to-text built into individual apps works a little differently in each one. A dedicated tool behaves the same everywhere.
One shortcut, everywhere
The same key press starts dictation in your email, your CRM, your browser and your chat. Learn it once — nothing to relearn per app.
Your volume drops to zero by itself (optional)
Your speaker volume goes to zero the moment you start recording and returns to its previous level when you are done.
Twenty minutes of recording, uninterrupted
Built-in dictation tends to cut out after a short pause. WritingAnt keeps listening for up to 20 minutes, so you can stop, think, and carry on at your own pace.
Think out loud. It types.
Saying it is easy — the words follow your thoughts. Typing is work: your head forms the sentence while your fingers hunt the keys.
20+ writing styles, informal to formal — your spoken thoughts rewritten to read like you typed them with care.
Mean every word? One style only tidies — fillers like “uhm” and “kinda”, commas, capitals — nothing rephrased.
- European GDPR
- Swiss FADP
- EU AI Act
- Your voice recording never leaves your computer. If anything is sent to the cloud at all, it is only data in text form, depending on the writing style.
- The Basic writing style runs fully on your device: voice recording and text never leave it.
- All other 20+ writing styles send only your text to the cloud for polishing, never your voice recording. Your text is never used to train AI models.
You choose where that text is processed: your speech-to-text results and the text you highlight for editing. The choice is made in the app, no enterprise plan needed.
Text processing optionAGlobal, including the United States- GPT AI models hosted by OpenAI.
- Reached through our cloud infrastructure on Microsoft Azure.
- The best AI models in terms of power and speed.
Text processing optionBSwitzerland & EuropeZero cloud data retention- Open-source AI models hosted by PHOENIQS in Switzerland.
- Reached through our service hosted on Scaleway in France (EU).
- Zero data retention in the cloud: the text you dictate or select for editing, and the result, are never logged or stored in the cloud. Keeping results in the text history on your own computer is a separate setting you control.
- No US-controlled provider in the chain, so immune to extraterritorial regulation such as the US CLOUD Act.
- Almost as accurate and fast as the global option.
Both options comply with the GDPR, the Swiss FADP and the EU AI Act.
- Enterprise: connect your own cloud AI. The polishing step then runs in your organisation's own AI environment, in the region you choose, under the terms of your own AI contract. Contact sales
Full list of sub-processors, processing locations and safeguards: Data Privacy PolicyData Processing Agreement (DPA)